Overview of diabetic nephropathy

INTRODUCTION Diabetic nephropathy occurs in type 1 (formerly called insulin-

dependent or juvenile onset) and type 2 (formerly called non-insulin-dependent or adult
onset) diabetes mellitus, and in other secondary forms of diabetes mellitus, for example
after pancreatitis or pancreatectomy, in which duration of diabetes is long-enough and level
of glycemia high enough to result in complications. (See "Classification of diabetes mellitus
and genetic diabetic syndromes".)
The following data concerning the epidemiology of renal disease may not represent the
current clinical course of the disease. Some of the evidence was obtained before the
availability of data supporting the efficacy of tight glycemic control, aggressive blood
pressure and lipid control, and the specific benefit of angiotensin-converting enzyme (ACE)
inhibitors or angiotensin II receptor blockers (ARBs).
This topic provides an overview of diabetic nephropathy; treatment issues are discussed
separately. (See "Treatment of diabetic nephropathy".)
CLINICAL FEATURES Diabetic kidney disease is defined by characteristic structural and
functional changes. The predominant structural changes include mesangial expansion,
glomerular basement membrane thickening, and glomerular sclerosis. (See 'Pathology'
below.)
The major clinical manifestations of diabetic nephropathy are albuminuria, less often
hematuria, and, in many patients, progressive chronic kidney disease, which can be slowed
or prevented with optimal therapy. (See "Treatment of diabetic nephropathy".)
Albuminuria Functional characteristics include hyperfiltration in very early disease
(primarily in type 1 diabetes); moderately increased albuminuria, formerly called
"microalbuminuria" (defined as urinary albumin excretion between 30 and 300 mg/day or
between 30 and 300 mg/g creatinine on a random urine sample); and severely increased
albuminuria, formerly called "macroalbuminuria" (defined as urinary albumin excretion
above 300 mg/day or above 300 mg/g creatinine on a random urine sample). Moderately
increased albuminuria precedes the development of severely increased albuminuria and is
considered to be a finding that predicts high risk for future nephropathy. The onset of
severely increased albuminuria is, in the absence of effective therapy, often followed by a
slowly progressive decline in glomerular filtration rate (GFR) and, given enough time, end-
stage renal disease [1]. Issues related to moderately increased albuminuria in patients with
diabetes are discussed elsewhere. (See "Moderately increased albuminuria
(microalbuminuria) in type 1 diabetes mellitus" and "Moderately increased albuminuria
(microalbuminuria) in type 2 diabetes mellitus".)
Progressive disease with normal or high-normal albuminuria For reasons that are
not understood, the degree of albuminuria is not necessarily linked to disease progression in
patients with diabetic nephropathy associated with either type 1 or type 2 diabetes [2-4].
This was illustrated in a report of 79 patients with type 1 diabetes from the Joslin Kidney
Study who were followed for a mean of 12 years after the onset of moderately increased
albuminuria [2]. Twenty-three of the patients progressed to advanced disease (GFR less
than 60 mL/min). Among these patients, 11 had either stable moderately increased
albuminuria or regression to normal albuminuria. The other 12 patients developed
proteinuria that usually accompanied but did not precede the decline in GFR. The rate of
loss of GFR was lower in patients with regression to normal albuminuria compared with
patients with stable moderately increased albuminuria or progression to severely increased
albuminuria [5]. Patients who progressed to severely increased albuminuria had the highest
rate of loss of GFR [5], and all six patients who developed end-stage renal disease had
progressed to severely increased albuminuria [2].
The factor(s) responsible for progressive GFR decline in nonalbuminuric diabetic
nephropathy are not known. One possibility is intrarenal vascular disease. However, in a
study of patients with type 2 diabetes and a GFR below 60 mL/min, intrarenal vascular
resistance on renal duplex scan was elevated to a similar degree in patients with and
without proteinuria [4]. In addition, nonalbuminuric progression occurs in type 1 diabetes, a
population that is generally much younger than those with type 2 diabetes and less likely to
have intrarenal vascular disease.
Hematuria The urine sediment in diabetic nephropathy is usually bland, but microscopic
hematuria can occur as it can in any form of glomerular disease, including disorders such as
membranous nephropathy that are not associated with glomerulonephritis. This is an
important issue in diabetic nephropathy since nondiabetic renal disease, either alone or with
diabetic nephropathy, is occasionally seen in patients with diabetes. (See 'Nondiabetic renal
disease' below.)
The prevalence of hematuria in patients with diabetic nephropathy and whether it is due to
diabetic nephropathy or a concurrent renal disease has been evaluated in a few studies [6-
8]. In a report from Japan of 154 patients with diabetes mellitus, glomerular hematuria was
observed in 26 (17 percent), 10 of whom had IgA nephropathy and one of whom had
membranous nephropathy [6]. The remaining 15 patients with hematuria, compared with
the 128 patients without hematuria, had more severe diabetic nephropathy on renal biopsy,
as manifested by advanced diffuse glomerular lesions and by interstitial lesions, and by a
significantly higher serum creatinine concentration.
Red blood cell casts have also been described in patients with diabetic nephropathy [7,8].
The clinical significance of this finding was evaluated in a study of eight patients with
hematuria, red blood cell casts, and known diabetes for 6 to 18 years [7]. Renal biopsy,
including immunofluorescence and electron microscopy, revealed glomerulonephritis in three
(postinfectious in two and IgA nephropathy in one). The other five patients had only diabetic
nephropathy.
EPIDEMIOLOGY
Type 1 diabetes The epidemiology of diabetic nephropathy has been best studied in
patients with type 1 disease since the time of clinical onset is usually known. Approximately
20 to 30 percent will have moderately increased albuminuria, formerly called
"microalbuminuria," after a mean duration of diabetes of 15 years [9-11]. Less than half of
these patients will progress to overt nephropathy; moderately increased albuminuria may
regress or remain stable in a substantial proportion, probably related to glycemic and blood
pressure control with angiotensin-converting enzyme (ACE) inhibitors and angiotensin
receptor blockers (ARBs). (See "Moderately increased albuminuria (microalbuminuria) in
type 1 diabetes mellitus", section on 'Prevalence' and "Moderately increased albuminuria
(microalbuminuria) in type 1 diabetes mellitus", section on 'Regression to
normoalbuminuria' and "Treatment of hypertension in patients with diabetes mellitus".)
Before it was commonplace for patients with type 1 diabetes to be managed with intensive
glycemic control and with angiotensin inhibition for blood pressure control, the incidence of
developing overt nephropathy (eg, a persistently positive urine dipstick for protein, or urine
albumin excretion >300 mg per day) and end-stage renal disease (ESRD) was high [9,12-
17]. As examples, the incidence of overt nephropathy was 25 to 45 percent, and the
incidence of ESRD was 4 to 17 percent at 20 years and 16 percent at 30 years of being
diagnosed with diabetes. In addition, after so-called severely increased albuminuria
developed (>300 mg albuminuria per day, formerly called "macroalbuminuria," also called
overt nephropathy or clinical grade proteinuria), the majority of patients progressed to
ESRD.
However, with intensive glycemic control and the use of ACE inhibitors and ARBs to control
blood pressure, patients diagnosed with type 1 diabetes may now have lower rates of overt
nephropathy and ESRD. The following illustrate the range of findings [14,18-21]:

A study from Sweden noted a dramatic reduction in clinically evident diabetic nephropathy
to 8.9 percent at 25 years, a presumed reflection of better glycemic control [18]. The
average hemoglobin A1c concentration in the later part of this follow-up period was 7.0
percent; patients without overt proteinuria had a lower hemoglobin A1c concentration than
those with proteinuria (7.1 versus 8.1 percent). A similar decline in incidence was not noted
in another report in which this degree of glycemic control was not achieved [20].

One study from Finland evaluated the long-term outcomes of 20,005 patients over the
period of 1965 to 1999 [19]. During the median follow-up period of 17 years (maximum of
37 years), progression to ESRD occurred in only 632 patients, with the cumulative incidence
being 2.2 and 7.8 percent at 20 and 30 years, respectively. In addition, the rate of ESRD
was relatively lower among patient cohorts diagnosed at later time points, and the incidence
of ESRD was lowest among those diagnosed prior to the age of 5 years.

In the DCCT/EDIC type 1 diabetes cohort, less than 2 percent (10 of 711) of the intensively
treated subjects developed renal insufficiency (serum creatinine >2.0 mg/dL or renal
replacement therapy) over an average of 30 years of diabetes duration [14].

In addition, severely increased albuminuria (urine albumin excretion >300 mg per day) no
longer heralds an inexorable progression to ESRD. Rather, albuminuria may regress in many
if not most of these patients, and rates of ESRD may be less than 20 percent at 10 years. In
the DCCT/EDIC type 1 diabetes cohort, for example, 123 patients developed severely
increased albuminuria that was persistent at two consecutive study visits [22]. During the
subsequent 10 years, 58 percent of these patients regressed to <300 mg of albuminuria per
day, and 12 percent regressed to <30 mg per day. The 10-year rates of reduced estimated
glomerular filtration rate (eGFR, <60 mL/min/1.73 m 2) and ESRD in the subjects with
severely increased albuminuria were 34 and 18 percent, respectively. Lower HbA1c and
blood pressure values were associated with a greater frequency of albuminuria regression
and a lower incidence of reduced eGFR and ESRD. Thus, even patients who develop overt
nephropathy can avoid progressive renal impairment with modern therapies.
The onset of overt nephropathy in type 1 diabetes is typically between 10 and 15 years after
the onset of the disease. Those patients who have no proteinuria after 20 to 25 years have
a risk of developing overt renal disease of only about 1 percent per year, at least among
Caucasians [12,21]. Thus, an alternative cause of renal disease should be considered in a
diabetic patient who develops new proteinuric kidney disease more than 20 years after the
diagnosis of type 1 diabetes.
Type 2 diabetes In Caucasians, the prevalence of progressive renal disease has
generally been lower in type 2 diabetes than in type 1 disease [23]. However, this
observation may be a function of the usually later-onset disease and shorter-duration
"exposure" in type 2 than type 1 diabetes, and may not apply to all groups with type 2
diabetes, some of whom have had a more ominous renal prognosis. In a 1990s study, for
example, nephropathy developed in up to 50 percent of diabetic Pima Indians at 20 years,
with 15 percent having progressed to ESRD by this time [24]. As previously described,
however, the use of modern therapies lowers the incidence of ESRD, even in groups at
extremely high risk such as the Pima Indians. In a subsequent study, for example, the
incidence of diabetic ESRD was noted to have declined significantly from the period 1991-
1994 to the period 1999-2002 (32 to 15 cases per 1000 patient-years, respectively) [25].
Data suggest that the renal risk is currently equivalent in the two types of diabetes.
Evidence in support of this hypothesis includes the observations in one report that the time
to proteinuria from the onset of diabetes and the time to ESRD from the onset of proteinuria
were similar in type 1 and type 2 disease [26,27].
Some of the most robust data relating to the development of diabetic nephropathy in a
population of predominantly white patients with type 2 was reported from the United
Kingdom Prospective Diabetes Study (UKPDS) [28]. The UKPDS was designed to compare
the efficacy of different treatment regimens (diet, oral hypoglycemics, insulin,
antihypertensive agents, varying blood pressure goals, and other interventions) on glycemic
control and the complications of diabetes (including renal failure) in newly diagnosed
patients with type 2 diabetes. The details of this study are described separately. (See
"Glycemic control and vascular complications in type 2 diabetes mellitus" and "Treatment of
diabetic nephropathy" and "Treatment of hypertension in patients with diabetes mellitus".)
With respect to the development and progression of nephropathy among almost 5100 type
2 diabetic patients enrolled in UKPDS, the following results were reported [28]:

At 10 years following diagnosis, the prevalence of moderately increased albuminuria,

severely increased albuminuria, and either an elevated plasma creatinine concentration
(defined as 175 micromol/L [2.0 mg/dL]) or requirement for renal replacement therapy
was 25, 5, and 0.8 percent, respectively.

The yearly rate of progression from diagnosis to moderately increased albuminuria, from
moderately increased albuminuria to severely increased albuminuria, and from severely
increased albuminuria to an elevated plasma creatinine concentration or renal replacement
therapy was 2.0, 2.8, and 2.3 percent.

Based upon a statistical model, an estimation of the median time spent in each stage
without progression to another nephropathy stage was 19, 11, and 10 years for those with
no nephropathy, moderately increased albuminuria, and severely increased albuminuria,
respectively. Among those with an elevated plasma creatinine concentration (2.0 mg/dL
[175 micromol/L]), renal replacement therapy was required after a median period of only
2.5 years. This rate of progression is higher than seen in other studies probably because of
two factors: appropriate therapy (eg, angiotensin inhibition and rigorous blood pressure
control) was not taken into account; and most of the patients had more advanced renal
insufficiency as defined by serum creatinine concentration 2.0 mg/dL [175 micromol/L].

As with type 1 diabetes, some patients with moderately increased albuminuria due to type 2
diabetes, particularly those with good glycemic control, experience regression of moderately
increased albuminuria [29].
PATHOLOGY Pathologic abnormalities are noted in patients with long-standing diabetes
mellitus before the onset of moderately increased albuminuria, formerly called
"microalbuminuria." There are three major histologic changes in the glomeruli in diabetic
nephropathy: mesangial expansion; glomerular basement membrane thickening; and
glomerular sclerosis [30,31].
The last abnormality, which may have a nodular appearance (the Kimmelstiel-Wilson lesion),
is often associated with hyaline deposits in the glomerular arterioles (reflecting the
insinuation of plasma proteins such as fibrin, albumin, immunoglobulins, and complement
into the vascular wall) (picture 1A-C) [30,32]. These different histologic patterns appear to
have similar prognostic significance [33].
The mesangial expansion and glomerulosclerosis do not always develop in parallel,
suggesting that they may have somewhat different underlying pathogenesis [32].
Renal Pathology Society classification A classification of type 1 and type 2 diabetic
nephropathy was developed by the research committee of the Renal Pathology Society [32].
Four classes of glomerular lesions were defined:

Class I: Isolated glomerular basement membrane thickening. Basement membranes are

greater than 430 nm in males older than age 9 and 395 nm in females. There is no evidence
of mesangial expansion, increased mesangial matrix, or global glomerulosclerosis involving
>50 percent of glomeruli.

Class II: Mild (class IIa) or severe (class IIb) mesangial expansion. A lesion is considered
severe if areas of expansion larger than the mean area of a capillary lumen are present in
>25 percent of the total mesangium.

Class III: At least one Kimmelstiel-Wilson lesion (nodular intercapillary glomerulosclerosis)

is observed on biopsy and there is <50 percent global glomerulosclerosis.

Class IV: Advanced diabetic sclerosis. There is >50 percent global glomerulosclerosis that
attributable to diabetic nephropathy.

The severities of interstitial and vascular lesions were also assigned scores:

A score of 0 was assigned if the interstitium had no areas of interstitial fibrosis and tubular
atrophy (IFTA); scores of 1, 2, or 3 were assigned if areas of IFTA <25, 25 to 50 or >50
percent, respectively.

A score of 0 was assigned if no T lymphocytes or macrophage infiltrate was present. Scores

of 1 or 2 were assigned if infiltrate was limited to the area surrounding atrophic tubules, or
if infiltrate was not limited, respectively.

Scores of 0, 1, or 2 were assigned if no arteriolar hyalinosis, one arteriole, or more than one
arteriole with hyalinosis was present. In addition, the most severely affected arteriole was
assigned a score of 0, 1, or 2 if there was no intimal thickening, intimal thickening <
thickness of media or intimal thickening > thickness of the media.

A pilot study in which five pathologists independently classified 25 biopsies demonstrated

good interobserver reproducibility (intraclass correlation coefficient [kappa statistic] of
0.84). Although this classification scheme is helpful to pathologists, its clinical utility is
unknown.
Other causes of nodular glomerulosclerosis Causes of nodular glomerulosclerosis
other than diabetic nephropathy include [33]:

Dysproteinemias such as amyloidosis and monoclonal immunoglobulin deposition diseases

(MIDD), mostly kappa light chain deposition disease. (See "Renal amyloidosis".)

Organized glomerular deposition diseases, fibrillary and immunotactoid glomerulonephritis,

fibronectin glomerulopathy, and collagen III glomerulopathy. (See "Glomerular diseases due
to nonamyloid fibrillar deposits" and "Fibronectin glomerulopathy".)

Chronic hypoxic or ischemic conditions, such as cyanotic congenital heart disease,

Takayasu's arteritis with renal artery stenosis, or cystic fibrosis [34].

Chronic membranoproliferative glomerulonephritis (type I). (See "Clinical presentation,

classification, and causes of membranoproliferative glomerulonephritis".)

Idiopathic nodular glomerulosclerosis, which is frequently associated with smoking,
hypertension and metabolic syndrome, but without overt diabetes mellitus [33,35,36].

Many of these conditions will be identified by characteristic findings observed by

immunofluorescence or electron microscopy.
PATHOGENESIS There appear to be different pathogenetic processes leading to the
pathologic mechanisms in diabetic nephropathy. Glomerulosclerosis, for example, may result
from intraglomerular hypertension induced by renal vasodilatation, or from ischemic injury
induced by hyaline narrowing of the vessels supplying the glomeruli [37].
Glomerular hyperfiltration The role of glomerular hypertension and hyperfiltration in
diabetic nephropathy is reinforced by the apparent benefits of blockade of the renin-
angiotensin system. Antagonizing the profibrotic effects of angiotensin II may also be a
significant factor in benefits observed with these agents [38]. Support for such profibrotic
elements underlying diabetic nephropathy is provided by findings in an animal model of
diabetic nephropathy [39]. Transient renin-angiotensin system blockade (for seven weeks)
in prediabetic rats reduced proteinuria and improved glomerular structure. (See 'Glomerular
filtration rate' below and "Mechanisms of glomerular hyperfiltration in diabetes mellitus".)
Hyperglycemia and AGEs Hyperglycemia may directly induce mesangial expansion and
injury, perhaps in part via increased matrix production or glycation of matrix proteins. In
vitro studies have demonstrated that hyperglycemia stimulates mesangial cell matrix
production [40,41], and mesangial cell apoptosis [42-44]. Mesangial cell expansion appears
to be mediated in part by an increase in the mesangial cell glucose concentration since
similar changes in mesangial function can be induced in a normal glucose milieu by
overexpression of glucose transporters, thereby enhancing glucose entry into the cells [41].
Glycation of tissue proteins also may contribute to the development of diabetic nephropathy
and other microvascular complications. In chronic hyperglycemia, some of the excess
glucose combines with free amino acids on circulating or tissue proteins. This nonenzymatic
process initially forms reversible early glycation products and later irreversible advanced
glycation end products (AGEs) (figure 1).
Circulating AGE levels are increased in diabetics, particularly those with renal insufficiency,
since AGEs are normally excreted in the urine [45]. The net effect is tissue accumulation of
AGEs, in part by crosslinking with collagen, which can contribute to the associated renal and
microvascular complications [46]. (See "Glycemic control and vascular complications in type
1 diabetes mellitus", section on 'Pathogenesis' and "Estimation of blood glucose control in
diabetes mellitus".)
Other proposed mechanisms by which hyperglycemia might promote the development of
diabetic nephropathy include activation of protein kinase C [47], and upregulation of
heparanase expression, as a decrease in cell surface heparan sulfate may contribute to
increased glomerular basement membrane permeability to albumin [48,49].
Prorenin Initial clinical studies in children and adolescents suggested that increased
plasma prorenin activity was a risk factor for the development of diabetic nephropathy
[50,51]. Prorenin binds to a specific tissue receptor that promotes activation of the mitogen-
activated protein kinases (MAPK) p44/p42A [52].
A possible pathogenic role for prorenin in the development of diabetic nephropathy was
suggested by an experimental model (streptozotocin diabetes in mice) in which prolonged
prorenin receptor blockade abolished MAPK activation and prevented the development of
nephropathy despite an unaltered increase in angiotensin II activity [53].
Cytokines Activation of cytokines, profibrotic elements, inflammation, and vascular
growth factors (vascular endothelial growth factor, VEGF) may be involved in the matrix
accumulation in diabetic nephropathy [54-69].
Hyperglycemia stimulates increased VEGF expression (a mediator of endothelial injury in
human diabetes) [55,56]. A potentially pathogenic role for VEGF in diabetic nephropathy is
supported by the observation that VEGF blockade improves albuminuria in an experimental
model of diabetic nephropathy [56,70].
A hyperglycemia-induced decrease in activated protein C contributes to the structural lesion
of diabetic nephropathy and worsens proteinuria in the streptozotocin model of diabetes in
mice [71]. The role of activated protein C in diabetic nephropathy in humans is not yet
known.
Hyperglycemia also increases the expression of transforming growth factor-beta (TGF-beta)
in the glomeruli and of matrix proteins specifically stimulated by this cytokine [54,65]. In
addition, diabetes is associated with decreased expression of renal bone morphogenic
protein-7 (BMP-7), which appears to counter the profibrogenic actions of TGF-beta [72].
The effect of some genetic factors may also be in part mediated by TGF-beta [73]. TGF-beta
may contribute to both the cellular hypertrophy and enhanced collagen synthesis that are
seen in diabetic nephropathy, illustrated by the examples below [73-77] (see 'Genetic
susceptibility' below):

The administration of an angiotensin-converting enzyme inhibitor to patients who have type

1 diabetes and nephropathy lowers serum concentrations of TGF-beta [75]. A correlation
has been found between decreased TGF-beta levels and renoprotection, as determined by
changes in the glomerular filtration rate over time.

The combination of an anti-TGF-beta antibody plus an ACE inhibitor completely normalized

proteinuria in rats with diabetic nephropathy; proteinuria was only partially lessened with an
ACE inhibitor alone [76]. Glomerulosclerosis and tubulointerstitial injury were also improved.

The administration of hepatocyte growth factor, which specifically blocks the profibrotic
actions of TGF-beta, ameliorates diabetic nephropathy in mice [77].

Nephrin expression The renal expression of nephrin may be impaired in diabetic

nephropathy. Congenital mutations in nephrin, a transmembrane protein expressed by
podocytes, result in severe congenital nephrotic syndrome of the Finnish type. (See
"Congenital and infantile nephrotic syndrome".)
When compared with nondiabetic patients with minimal change nephropathy and controls,
patients with diabetic nephropathy had markedly lower renal nephrin expression and fewer
electron dense slit diaphragms [78]. By comparison, the expression of two other important
podocyte/slit diaphragm proteins, podocin and CD2AP, was similar among the three groups.
Similar decreases in renal nephrin expression have been reported in other studies of
diabetic nephropathy [79,80].
Impaired podocyte-specific insulin signaling Defects in podocyte-specific insulin
signaling may contribute to diabetic nephropathy. Mouse models have been generated in
which the gene encoding the insulin receptor is deleted in a podocyte-specific manner [81].
In the absence of hyperglycemia, affected mice develop albuminuria, effacement of foot
processes, apoptosis, glomerular basement membrane thickening, accumulation of
mesangial matrix and glomerulosclerosis. Activation of the insulin receptor appears to
trigger remodeling of the actin cytoskeleton through the mitogen-activated protein kinase
42/44 (MAPK) and phosphatidylinositol 3 (PI3) kinase signaling pathways, suggesting a
possible mechanism of proteinuria. The podocyte insulin receptor may provide a target for
agents that prevent proteinuria and/or development and progression of diabetic
nephropathy [82,83].
RISK FACTORS Studies in patients who have or do not have clinically evident diabetic
nephropathy have identified a number of factors as being associated with increased risk of
renal involvement [30,84-86].
Genetic susceptibility Genetic susceptibility may be an important determinant of both
the incidence and severity of diabetic nephropathy [31,47,87]. The likelihood of developing
diabetic nephropathy is markedly increased in patients with a diabetic sibling or parent who
has diabetic nephropathy; these observations have been made in both type 1 and type 2
diabetes [88-90]. One report, for example, evaluated Pima Indian families in which two
successive generations had type 2 diabetes [89]. The likelihood of the offspring developing
overt proteinuria was 14 percent if neither parent had proteinuria, 23 percent if one parent
had proteinuria, and 46 percent if both parents had proteinuria.
The familial increase in risk cannot be explained by the duration of diabetes, hypertension,
or the degree of glycemic control, although a genetic predisposition to abnormal sodium
handling and hypertension may be important [88]. Genetics also may contribute to the
influence of race on diabetic nephropathy (see 'Race' below).
Many studies have examined the role of the angiotensin-converting enzyme (ACE) gene
genotype as a potential genetic risk factor with conflicting data:

In patients with type 2 diabetes, the DD polymorphism has been associated with an
increased risk for the development of diabetic nephropathy, more severe proteinuria,
greater likelihood of progressive renal failure, and mortality on dialysis [91-93].

A critical review of nineteen studies examining a possible link between the ACE gene
genotype and diabetic nephropathy failed to confirm an association among Caucasians with
either type 1 or type 2 diabetes, but could not exclude a possible association in Asians [94].
Unfortunately, due to poor methodology, no definitive conclusions could be drawn.

An analysis of over 1,000 Caucasian patients with type 1 diabetes found a strong correlation
between genetic variation in the ACE gene and the development of nephropathy [95].
However, a family-based European study found no correlation between diabetic nephropathy
in type 1 diabetics and certain ACE polymorphisms [96].

Another implicated genetic factor is the angiotensin-II type 2 receptor gene (AT2) on the X-
chromosome. Male patients with type 1 diabetes and the AA haplotype of the AT2 gene had
a lower glomerular filtration rate than those with the GT haplotype [97]. A similar
association was not observed among females.
An enhanced risk may also be due to inheritance of one allele of the aldose reductase gene,
the rate-limiting enzyme for the polyol pathway. In a controlled study of patients with type 1
diabetes, homozygosity for the Z-2 allele was significantly associated with an odds ratio of
5.25 for the presence of nephropathy [98]. (See "Aldose reductase inhibitors in the
prevention of diabetic complications".)
Variants of the PKCb1 gene have been associated with kidney disease in Chinese patients
with type 2 diabetes [99]. Four common single-nucleotide polymorphisms were
independently associated with end-stage renal disease in two large cohorts of Chinese
patients.
The Genetics of Kidneys in Diabetes Study (GoKinD) has assembled a cohort of individuals
and families with type 1 diabetes with and without kidney disease to facilitate examination
of potential genetic factors predisposing to diabetic nephropathy [100].
Age The impact of age at onset of diabetes on the risk of developing nephropathy and
end-stage renal disease is unclear. As an example, among patients with type 2 diabetes,
increasing age, along with increasing duration of diabetes, has been associated with an
increased risk for developing albuminuria in Australia [86]. In contrast, in a population-
based study of 1856 Pima Indians with type 2 diabetes, patients who developed diabetes
prior to age 20 had a higher risk of progressing to end-stage renal disease (25 versus 5 per
1000 patient years at risk) [101].
For type 1 diabetes, the risk of developing ESRD is very low for patients diagnosed prior to
age 5; at older ages, the relationship of age to progression to ESRD is uncertain [19,102].
Blood pressure Prospective studies have noted an association between the subsequent
development of diabetic nephropathy and higher systemic pressures. The pathogenesis of
hypertension, the goal blood pressure, and the choice of antihypertensive drugs in patients
with diabetes are discussed elsewhere. (See "Treatment of hypertension in patients with
diabetes mellitus".)
Glomerular filtration rate Approximately one-half of patients with type 1 diabetes of
less than five years duration have an elevated glomerular filtration rate (GFR) that is 25 to
50 percent above normal. (See "Mechanisms of glomerular hyperfiltration in diabetes
mellitus".)
Those patients with glomerular hyperfiltration appear to be at increased risk for diabetic
renal disease [85,103]. This is particularly true for overt nephropathy if the initial GFR is
above 150 mL/min; in comparison, lesser degrees of hyperfiltration may have a slower
course, with a lesser risk for moderately increased albuminuria, formerly called
"microalbuminuria." In one prospective study, for example, patients with type 1 diabetes
and a GFR above 125 mL/min had a risk of developing moderately increased albuminuria
within 8 years of approximately 50 percent versus only 5 percent in patients with a lower
GFR that was similar to that seen in nondiabetics [103].
The glomerular hyperfiltration in type 1 diabetics is typically associated with glomerular
hypertrophy and increased renal size [104]. The association between these hemodynamic
and structural changes and the development of diabetic nephropathy may be related both to
intraglomerular hypertension (which drives the hyperfiltration) and to glomerular
hypertrophy (which also increases wall stress). Therapy aimed at reversing these changes
with aggressive control of the plasma glucose concentration early in the course of the
disease [104], dietary protein restriction, and antihypertensive therapymay slow the rate
of progression of the renal disease. (See "Treatment of diabetic nephropathy".)
The findings in type 2 diabetes are somewhat different. Up to 45 percent of affected
patients initially have a GFR that is more than 2 standard deviations above age-matched
nondiabetic and obese controls [105,106]. However, the degree of hyperfiltration (averaging
117 to 133 mL/min) is less than that seen in type 1 diabetics. Type 2 diabetics are also
older and more likely to have arteriosclerotic vascular disease, which will limit increases in
both glomerular filtration and glomerular size [107].
The potential importance of intraglomerular hypertension in the pathogenesis of diabetic
nephropathy may explain why systemic hypertension is an important risk factor for the
development of diabetic nephropathy [108]. Studies in experimental animals suggest that
the diabetic state is associated with impaired renal autoregulation. As a result, raising the
systemic pressure does not produce the expected afferent arteriolar vasoconstriction that
would minimize transmission of the elevated pressure to the glomerulus [109].
Glycemic control Diabetic nephropathy is more likely to develop in patients with worse
glycemic control (higher HbA1c levels). This has been well established in controlled clinical
trials and is discussed in detail separately. (See "Glycemic control and vascular
complications in type 1 diabetes mellitus" and "Glycemic control and vascular complications
in type 2 diabetes mellitus".)
Race The incidence and severity of diabetic nephropathy are increased in blacks (3- to 6-
fold compared to Caucasians), Mexican-Americans, and Pima Indians with type 2 diabetes
[24,110,111]. This observation in such genetically disparate populations suggests a primary
role for socioeconomic factors, such as diet, poor control of hyperglycemia, hypertension,
and obesity [112].
As an example, there appears to be an important association between hypertension and
disease progression in black patients with type 2 diabetes. A cross-sectional study found
that the GFR was normal in patients who were normotensive; in comparison, hypertension
was associated with a decline in renal function, particularly if it developed after the onset of
diabetes and the patient had been diabetic for more than 10 years [113]. It is not clear,
however, if the hypertension worsened the renal disease or was simply a marker for more
severe renal involvement. (Blacks are also at increased risk for some other renal diseases,
particularly hypertensive nephrosclerosis, (see "Hypertensive complications in blacks").)
However, the importance of genetic influences in the racial propensity to diabetic
nephropathy cannot be excluded. Even when adjustments are made for the increased
incidence of hypertension and lower socioeconomic status in blacks, there still appears to be
as much as a 4.8-fold increase in the risk of end-stage renal disease due to diabetic
nephropathy in blacks [110]. This appears to occur only in type 2 diabetes, with no increase
in risk seen with type 1 diabetes.
Pima Indians, on the other hand, have larger glomeruli than Caucasians, a finding that may
represent a specific genetic trait [114]. This increase in glomerular size, via the mechanism
described above, could lead to enhanced susceptibility to diabetes-induced glomerular
injury. One manifestation of this increased risk is the observation that diabetic Pima Indians
with a known duration of disease of less than 3 years already have evidence of glomerular
dysfunction (increased albumin excretion due to impaired glomerular size-selectivity)
[24,115]. These patients also have a higher GFR (140 versus 122 mL/min) when compared
to matched patients without diabetes [115].
Obesity A high body mass index (BMI) has been associated with an increased risk of
chronic kidney disease among patients with diabetes [86,116-118]. In addition, diet and
weight loss may reduce albuminuria and improve kidney function among patients with
diabetes [119,120]. However, the contribution of obesity (or weight loss) to the risk of
nephropathy (independent of diabetes and glycemic control) has not been convincingly
demonstrated in these studies.
Smoking Smoking is associated with a variety of adverse effects in patients with
diabetes. This includes evidence of increases in albuminuria and the risk of end-stage renal
disease and of decreased survival once dialysis is begun. Some of these effects may be
related to worse glycemic control. The supporting data are presented separately.
Oral contraceptives An initial report suggested a link between oral contraceptive use
and the risk of diabetic nephropathy [121].
Predictive capability Although each of the above factors increases the risk of
developing diabetic nephropathy, none is as yet sufficiently predictive in the individual
patient. The earliest detectable sign of diabetic nephropathy is moderately increased
albuminuria, which is associated with a substantial risk of progressive renal damage [85].
(See "Moderately increased albuminuria (microalbuminuria) in type 1 diabetes mellitus".)
Severely increased albuminuria, formerly called "macroalbuminuria," is far more predictive
of the future development of severe nephropathy with decreased GFR.
RELATION BETWEEN DIABETIC NEPHROPATHY AND RETINOPATHY Patients with
nephropathy and type 1 diabetes almost always have other signs of diabetic microvascular
disease, such as retinopathy and neuropathy [9,12]. The retinopathy is easy to detect
clinically, and typically precedes the onset of overt nephropathy in these patients. The
converse is not true. Only a minority of patients with advanced retinopathy have histologic
changes in the glomeruli and increased protein excretion that is at least in the moderately
increased albuminuria range, and most have little or no renal disease as assessed by renal
biopsy and protein excretion [122,123].
The relationship between diabetic nephropathy and retinopathy is less predictable in type 2
diabetes. In one study of 35 patients with diabetes and significant proteinuria (>300
mg/day), 27 (77 percent) were found to have diabetic nephropathy by biopsy [124].
Diabetic retinopathy was present in 15 of the 27 (56 percent), and in none of the eight
individuals without diabetic nephropathy, thereby resulting in a positive and negative
predictive value of 100 and 40 percent, respectively. Further analysis of some of these
patients plus additional type 2 diabetics with proteinuria found that, among those without
retinopathy, approximately 30 percent did not have diabetic nephropathy upon renal biopsy
[125].
Thus, type 2 diabetics with marked proteinuria and retinopathy most likely have diabetic
nephropathy, while those without retinopathy have a high frequency of nondiabetic
glomerular disease [126]. Data from the third National Health and Nutrition Examination
Survey suggest that 30 percent of type 2 diabetics with renal insufficiency have nondiabetic
renal disease, as inferred by the absence of albuminuria and retinopathy in this population
[127].
Some insight into the correlation between nephropathy and retinopathy was provided by a
study in which 36 patients with type 2 diabetes and renal involvement underwent renal
biopsy [128]. Seventeen biopsies displayed diabetic glomerulosclerosis with Kimmelstiel-
Wilson nodules, while 15 revealed changes characteristic of diabetes (mesangial sclerosis)
but without classic nodules. Patients with and without nodules did not differ with respect to
duration of diabetes or degree of glycemic control. A close correlation was observed
between the presence of severe retinopathy and nodules on biopsy: six of seven patients
with proliferative retinopathy had such nodules, while seven of eight without retinopathy
had mesangial sclerosis. Overall, severe retinopathy was more closely associated with
nodules than with mesangial sclerosis. The reason for this association is unknown.
Blockade of the renin-angiotensin system slows the rate of progression of diabetic
nephropathy and may have a similar benefit in diabetic retinopathy. In a randomized trial of
285 patients with type 1 diabetes, normal blood pressure, and normoalbuminuria, therapy
with losartan or enalapril reduced the odds of retinopathy progression by 65 to 70 percent,
compared with placebo [129]. (See "Treatment of diabetic nephropathy".)
Based upon the correlation between retinopathy and nephropathy, the 2007 K/DOQI
Guidelines for diabetes and chronic kidney disease suggest that chronic kidney disease
should be attributed to diabetes in most patients with diabetes if moderately increased
albuminuria, formerly called "microalbuminuria," or proteinuria and diabetic retinopathy are
both present [130]. By comparison, other causes of CKD should be entertained if diabetic
retinopathy is absent. (See 'Nondiabetic renal disease' below.)
NONDIABETIC RENAL DISEASE Proteinuria and/or hematuria in diabetes mellitus is
occasionally due to a glomerular disease other than diabetic nephropathy. As examples,
membranous nephropathy, minimal change disease, IgA nephropathy, focal segmental
glomerulosclerosis, Henoch-Schnlein purpura (IgA vasculitis), thin basement membrane
disease, proliferative glomerulonephritis, collapsing glomerulopathy, and pauci-immune
crescentic glomerulonephritis have all been described [6,7,107,124,125,131-144]. Since
renal disease in the setting of diabetes is often ascribed to the diabetes, without further
diagnostic efforts, the coincidence of nondiabetic renal disease in persons with diabetes may
be underestimated [143].
With respect to membranous nephropathy, porcine insulin may be the inciting antigen in
some patients. This was suggested in a report of seven patients with diabetes who
developed membranous nephropathy [132]. Three of the patients had granular deposits of
insulin along the glomerular capillary wall, as detected by anti-porcine insulin antibodies,
along with IgG and complement. The nephrotic syndrome developed after the initiation of
porcine insulin in all three patients and, in two of these patients, the use of human insulin
rather than porcine insulin led to amelioration of the proteinuria. The other four patients had
idiopathic membranous nephropathy. (See "Causes and diagnosis of membranous
nephropathy".)
The major clinical clues suggesting the presence of nondiabetic glomerular disease are
[124,130,134,145,146]:

The onset of proteinuria less than five years from the documented onset of type 1 diabetes,
since the latent period for overt diabetic nephropathy is usually at least 10 to 15 years
[9,12]. The latent period is probably similar in patients with type 2 diabetes, but the time of
onset is often difficult to ascertain.

The acute onset of renal disease. Diabetic nephropathy is a slowly progressive disorder
characterized by increases in protein excretion and the serum creatinine concentration over
a period of years.

The presence of an active urine sediment containing red cells (particularly acanthocytes)
and cellular casts. However, hematuria and red cell casts can also be seen with diabetic
nephropathy alone [6-8,146]. (See 'Hematuria' above.)

In type 1 diabetes, the absence of diabetic retinopathy or neuropathy. In contrast, lack of

retinopathy in type 2 diabetes does not preclude diabetic nephropathy, which was absent in
12 of 27 patients with biopsy confirmed diabetic nephropathy in one study [124].

Signs and/or symptoms of another systemic disease [130].

Significant reduction in the glomerular filtration rate (>30 percent) within two to three
months of the administration of ACE inhibitors or angiotensin II receptor blockers [130].

The reported frequency of other renal diseases among patients with diabetes depends upon
multiple factors, including ethnicity, geographic location, and a given institution's biopsy
policy. The importance of biopsy policy was evaluated in a multicenter study in Italy of
biopsy results in 393 patients with type 2 diabetes [138]. The biopsy policy was either
restricted (in which the procedure was performed for the indications just mentioned and
some less stringent criteria) or unrestricted (in which the procedure was performed for
much less stringent criteria). The following findings were noted:

With the restricted biopsy policy, 29 percent had diabetic nephropathy alone, 38 percent
had another glomerular disease superimposed on diabetic nephropathy, and the remaining
patients had either nephrosclerosis or nondiabetic glomerular disease.

With the unrestricted policy, the rate of diabetic nephropathy alone was much higher (51
percent), 22 percent had another glomerular disease superimposed on diabetic
nephropathy, and the remaining patients had either nephrosclerosis or nondiabetic
glomerular disease.

The most common nondiabetic glomerular diseases were membranous nephropathy, IgA
nephropathy, postinfectious glomerulonephritis, and minimal change disease or focal
glomerulosclerosis. These disorders typically have a relatively rapid onset in contrast to the
slow progression over years from moderately increased albuminuria, formerly called
"microalbuminuria," to severely increased albuminuria, formerly called "macroalbuminuria,"
in diabetic nephropathy. (See 'Epidemiology' above.)

Nephrosclerosis In addition to nondiabetic glomerular diseases, renal insufficiency and

proteinuria may also be induced by other diseases, particularly arteriosclerotic vascular
disease (nephrosclerosis) in older type 2 diabetics [27,107,138,147]. This disorder cannot
usually be distinguished from diabetic nephropathy without performing a renal biopsy; this
is rarely necessary since making this distinction is of no clinical value. One potential clue
favoring the presence of nephrosclerosis is a rise in the plasma creatinine concentration,
due to interference with renal autoregulation, after institution of an angiotensin-converting
enzyme inhibitor to treat hypertension or to slow the rate of progression of the renal
disease. However, this is also consistent with renal artery stenosis. (See "Renal effects of
ACE inhibitors in hypertension" and "Treatment of diabetic nephropathy".)
INFORMATION FOR PATIENTS UpToDate offers two types of patient education
materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are
written in plain language, at the 5 th to 6th grade reading level, and they answer the four or
five key questions a patient might have about a given condition. These articles are best for
patients who want a general overview and who prefer short, easy-to-read materials. Beyond
the Basics patient education pieces are longer, more sophisticated, and more detailed. These
articles are written at the 10 th to 12th grade reading level and are best for patients who want
in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to
print or e-mail these topics to your patients. (You can also locate patient education articles
on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

Beyond the Basics topic (see "Patient education: Diabetic kidney disease (diabetic
nephropathy) (Beyond the Basics)")

SUMMARY

Diabetic nephropathy occurs in both type 1 and type 2 diabetes mellitus. (See 'Introduction'
above.)

The risk of progression of diabetic nephropathy has improved over the last several decades,
largely due to rigorous glycemic control, more aggressive blood pressure reduction and the
use of angiotensin-converting enzyme inhibitors. (See 'Epidemiology' above.)

The three major histologic changes in the glomeruli in diabetic nephropathy include
mesangial expansion; glomerular basement membrane thickening; and glomerular sclerosis.
Different histologic patterns have similar prognostic significance. (See 'Pathology' above.)

Pathogenetic processes that contribute to diabetic nephropathy in susceptible individuals

include glomerular hyperfiltration; hyperglycemia and the increased production of advanced
glycation end products, increased plasma prorenin activity; hypoxia-inflammation and the
activation of cytokines. (See 'Pathogenesis' above.)

Risk factors associated with diabetic nephropathy include family history of diabetes, black
race, Mexican-American or Pima Indian ancestry; higher systemic blood pressures, evidence
of hyperfiltration early in course of disease; poor glycemic control; smoking; and possibly
the use of oral contraceptives. Obesity and older age may also be risk factors. No one factor
is predictive in the individual patient. (See 'Risk factors' above.)

Patients with nephropathy and type 1 diabetes almost always have other signs of diabetic
microvascular disease, such as retinopathy and neuropathy. The relationship between
diabetic nephropathy and retinopathy is less predictable in type 2 diabetes. The 2007
K/DOQI Guidelines for diabetes and chronic kidney disease suggest that chronic kidney
disease should be attributed to diabetic nephropathies in most patients with diabetes if
albuminuria and diabetic retinopathy are both present; other causes of CKD should be
entertained if diabetic retinopathy is absent. (See 'Relation between diabetic nephropathy
and retinopathy' above.)

Albuminuria in diabetes mellitus is occasionally due to a glomerular disease other than

diabetic nephropathy, or to nephrosclerosis. The major clinical clues suggesting nondiabetic
glomerular disease include the onset of proteinuria less than five years from the
documented onset of type 1 diabetes; the acute onset of renal disease; the presence of an
active urine sediment containing red cells (particularly acanthocytes) and cellular casts;
signs and/or symptoms of another systemic disease; or significant reduction in the
glomerular filtration rate (>30 percent) within two to three months of the administration of
ACE inhibitors or angiotensin II receptor blockers. (See 'Nondiabetic renal disease' above
and 'Nephrosclerosis' above.)
Use of UpToDate is subject to the Subscription and License Agreement.
REFERENCES